Good designs, according to Norman, have visual cues as to how they are to be used. Norman uses the example of a door. When you approach a door, you immediately judge, consciously or not, how the door should be opened. Should it be pushed, pulled, or slid to one side or the other? Usually, you do not need to put much thought into the decision. When the door has a vertical handle, you assume that you need to pull to open it. If the door has a horizontal handle or no handle at all, you assume you need to push, and, as Norman explains, you push on the unsupported side. However, when the designer decides that he or she wants to trump clarity of use with beauty, some of these visual clues may disappear, as in the case of the doors in the European post office mentioned in the article.
Proper feedback is essential to a good design; types of feedback can be as simple as a door swinging open or technologically sophisticated as an audio confirmation to convey that a task has been successfully performed. Lack of feedback is one of the main reasons why the telephone systems discussed in Norman's article were poorly designed. You could push almost any button (and chances are that that button is unmarked) in an attempt to use one of the phone's many nonessential features, but you would be given no confirmation on whether or not the job was carried out.
Simplicity must be taken into account when attempting to create the "perfect" design, especially the simplicity of visual cues. This aspect of a design presents a challenge: where do you draw the lines between too simple and not simple enough? The doors to the European post office were too simple; there were not enough visual cues for Norman's friend to easily determine whether to push or pull. On the other hand, the example of the washing machine demonstrates how a design can be far from simple enough. The excessive amount of settings and functions overwhelmed the couple who bought the machine, and they resorted to memorizing one setting and ignoring the rest.
Norman used the example of a Mercedes Benz to illustrate a good relationship between controls and their outcomes in a good design. For instance, consider the controls used to adjust the position of the seat. The control itself was shaped like the seat; if you wanted to raise the front of the seat, you would tilt the front of the control up. The relationship is clear and obvious, requiring no instructions.
All of these examples illustrate Norman's main point: a good design cooperates with a person's natural approach to the object; it does not hinder it.
2. At home back in Dearborn, the oven my family owns is one of the most irritating appliances in the house. I have many complaints about it, such as the fact that despite the oven having two shelves, you can only use one if you want something cooked correctly. Another flaw is the combination of the timer and the temperature control of the oven itself. There is only one button labeled off/clear. Somehow this is supposed to apply to both the temperature and the timer. Ninety percent of the times I (attempt to) use it, I have discovered too late that I turned the entire oven off when I meant to only turn the timer off, and left the food uncooked or only halfway cooked, resulting in the pizza delivery guy knocking on the door half an hour later.
3. As far as the design of the iPod is concerned, Fadell and the designing team could have gone exactly by the principles discussed in Norman's article. The iPod has just the right amount of visual cues to hint to the user how to operate the device. The iPod obviously has effective feedback. If you don't hear music, you know you must have missed something. But with such effective mappings and such clear relationships between the controls and their outcomes, chances are that you'll get the feedback you are looking for every time.